Response times for non-emergency calls can have a huge impact on the public’s confidence in law enforcement. But due to recruitment and staff retention factors, as well as the increased complexity of the crimes reported, answer times are slowing down.
It’s a challenge that police forces have been looking to address, according to Rob Brind, a digital silver commander (a senior officer in the force’s Gold-Silver-Bronze command hierarchy) at Hampshire & Isle of White Constabulary, and Thames Valley Police.
“We’re able to maintain our emergency 999 call answer times, but we’d got to the point where our non-emergency 101 call rates were taking, on average, seven minutes to answer,” says Brind.
“To just handle a call now takes longer. Most crimes are digitally enabled, be they phishing scams or a stalking harassment case via social media…All this increased complexity of gathering information is usually done by the call handler, which has seen the number of calls drop from four-to-five an hour to two-to-three,” he explains.
Brind found that half of all calls coming into the police were incident descriptions after conducting analysis of the types of non-emergency demand that was coming in. Another 20% of demand was from callers looking for updates on reported crimes, while an additional 30% of calls were harder to predict.
“In these cases, we don’t really know what the demand looks like,” Brind admits.
According to the commander, tackling the 20% of people calling in for updates was critical – as this was the category which police could anticipate and therefore cater for more efficiently using automated technologies.
“We were looking for something that could either push automated messages to these victims of crime who were looking for updates or allow them to access messages or policing services in a different way that would keep them informed and engaged,” Brind explains.
After looking at a range of technologies that could assist the force in sensitively automating a crime victim’s journey, the force chose Salesforce’s Marketing Cloud solution.
“This allowed us to create a journey, and send automated messages, text messages or emails to those victims over a period of the crime lifecycle so. So, info about the crime number, the officer assigned to the case and crime prevention advice.
“Now, instead of the victim having to call us they’re getting that info straight away over a period of time. This information is coming either from our contact management system or our record management system.”
Since the force went live with the system last August, Brind says that the system has sent over 400,000 messages to victims of crime.
Thames Valley Police also worked with Salesforce to build a portal where victims can self-serve information and monitor the progress of the investigation.
“On the portal, there’s a chevron at the top that tells you what status your crime is at. Whether it’s being investigated, or whether stuff is out for forensics etc”
“We’ve also built a two-way messaging service between officers and members of the public. So instead of victims of crime having to phone up for an update or wait for the officer to contact them, they can now contact the officer and the officer can message back through the portal,” Brind explains.
Following the portal’s launch in February, it has already exchanged 8,000 messages between officers and members of the public. “We’ve also seen over 40,000 people access the portal over that period as well,” Brind adds.
The portal has also been designed to host useful information: from victim code of compliance; knowledge articles complied by the Home Office around support services; as well as a list of NGOs and specialists that victims of crime can speak with.
Brind estimates that since the technology was installed there has been a call reduction of 30,000 across Hampshire and Thames Valley in the last quarter.
“That’s a significant number of calls reduced. If that continues that’s 120,000 calls reduced to call centres which allows us to focus on other areas of demand – people who are victims or witnesses of crime and recording it and dealing with that in a timely manner.”
Brind adds that speed-to-answer times have reduced from seven minutes to three-to-four minutes over the last month or so. He enthuses: “We are giving members of the public a more efficient service because they are getting updates in a timelier way, and they are getting a choice now of how they get updated on what’s happening.”
The commander admits that not all sections of the public likes automated messages, but the aim is to address a significant portion of the demand.
“Around 90% of people are quite happy being pushed info, another 9% will want to self-serve because they need further info. And 1% will always want to call up and ask for more. If we can tackle a portion of that 90%, we are halfway there to dealing with a much better service,” he says.
Of course, there are more things on the list. The automated service does not yet work in other languages, but Brind adds “AI is getting better at doing that” and he’s hopeful that this is something they can add soon.
The force would also like to extend its automated service to other parts of the law enforcement system such as witness care.
“There’s an attrition rate of victims and witnesses not going to court because they haven’t been updated. Post-Covid those waits are longer because there’s a backlog in the courts.
“So, how do we deal with those witnesses and people better, so we get those cases to court? Pushing info or self-serving on a portal could support this process better,” he says.
Another challenge the force would like to address is the 30% of calls which are harder to predict. These are calls that neither fall under the category of ‘crime’ or ‘incident’ but might be related to healthcare, mental health services or the local authorities.
“Over the next 12 months, we want to look at how to get better at diverting those calls away from 101 to the right person at the right time.”
Thames Valley Police is currently working with AI-powered speech analytics software company Sprinklr to address this challenge and create a chatbot that can deal with multiple calls at any one time.
Brind is also looking at how the banking sector uses AI-powered bots, like NatWest’s Cora, to see how it can be applied to policing.
The commander adds that these bots would also help with crime reports made online.
“We use robotics to process crime reports and that info goes into our record management system. But those that can’t be processed by a robot need to be done manually. So how do we automate that process going forward? In a way using AI and running AI over the top…”
While budgets are tight, TI asks Brind what he’d like to do if he had a larger budget. Like many organisations, he replies that in an increasingly AI-driven world, a deep data cleanse would benefit the force.
He explains that much of the force’s data has been moved to a Microsoft Azure cloud, but tasks like sorting out structured data from unstructured data and releasing siloed data so that crime reports, can be brought together with contact data or HR data would be “hugely powerful”.
He adds: “I’d like to see greater investment in being able to pull together our data: how we cleanse it; use it to inform our long-, medium- and short-term decision-making.
“The vast majority of our budgets go on people so our expenditure on digital and technology is limited. But it would be great to bring together our datasets and our partnership data sets, bring that together and use it to inform our decision making.”
“That sounds simplistic but that’s all we want to do in police – be a bit more proactive around making sure we’ve got the right resources in the right teams with the right skills to do what we need to do in the future.”
Subscribe to our Editor's weekly newsletter
Newsletter
Share
